DescriptionKonstantin Khorenko
2008-12-11 11:25:06 UTC
General description: multipath does not work by default, dmesg shows following messages after multipath tool is executed:
...
kernel: device-mapper: table: 253:0: multipath: unknown hardware handler type
kernel: device-mapper: ioctl: error adding target to table
...
The bug was found under CentOS 5.2 nut i believe it's also valid for RHEL 5.2.
OS was installed by default with "linux mpath" option.
Hardware that triggers the problem:
lspci, lspci -n:
03:01.0 Fibre Channel: QLogic Corp. ISP2312-based 2Gb Fibre Channel to PCI-X HBA (rev 02)
03:01.0 0c04: 1077:2312 (rev 02)
# cat /proc/scsi/scsi
Attached devices:
...
Host: scsi1 Channel: 00 Id: 00 Lun: 00
Vendor: COMPAQ Model: HSV111 (C)COMPAQ Rev: 4007
Type: RAID ANSI SCSI revision: 05
...
If /etc/multipath.conf does not contain "hardware_handler" option (by default), multipath consider that hardware handler "hp_sw" should be used:
[root@localhost ~]# multipath -v2 -d
create: mpath0 (3600508b400011c130003000000fc0000) COMPAQ,HSV111 (C)COMPAQ
[size=100G][features=0][hwhandler=1 hp_sw]
\_ round-robin 0 [prio=4][undef]
\_ 0:0:0:1 sda 8:0 [undef][ready]
\_ 0:0:1:1 sdb 8:16 [undef][ready]
\_ 1:0:0:1 sdc 8:32 [undef][ready]
\_ 1:0:1:1 sdd 8:48 [undef][ready]
If we add correct "hardware_handler" option for the device - the problem is gone. So following solves the problem.
devices {
device {
vendor "HP|COMPAQ*"
product "HSV1*"
hardware_handler "1 hp-sw"
}
}
i think this bug is related to
https://bugzilla.redhat.com/show_bug.cgi?id=348951https://bugzilla.redhat.com/show_bug.cgi?id=226791
it's said the bug is fixed in device-mapper-multipath-0.4.7-17.el5 (in errata), but my problem triggers on device-mapper-multipath-0.4.7-17.el5!
[root@localhost ~]# cat /etc/redhat-release
CentOS release 5.2 (Final)
[root@localhost ~]# rpm -qa |grep mapper
device-mapper-event-1.02.24-1.el5
device-mapper-multipath-0.4.7-17.el5
device-mapper-1.02.24-1.el5
------------------------------------------
Next issue which might be related:
mkinitrd (mkinitrd-5.1.19.6-28) script does not put dm-hp-sw.ko module into initrd. I believe this is because of the previous issue.
To workaround this i've added following line to the /etc/modprobe.conf:
alias scsi_hostadapter2 dm-hp-sw
This is incorrect from the logic point of view, but this will give us needed result - dm-hp-sw.ko module will be added into initrd.
------------------------------------------
And the last one:
mkinitrd script does not handle "install" primitive for qla2xxx module in /etc/modprobe.conf (which was set by installer!) and thus qla2xxx.ko module is _not_ put into initrd.
There is already a bug for it https://bugzilla.redhat.com/show_bug.cgi?id=317811, but it was closed as won't fix.
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.
http://rhn.redhat.com/errata/RHBA-2010-0255.html
General description: multipath does not work by default, dmesg shows following messages after multipath tool is executed: ... kernel: device-mapper: table: 253:0: multipath: unknown hardware handler type kernel: device-mapper: ioctl: error adding target to table ... The bug was found under CentOS 5.2 nut i believe it's also valid for RHEL 5.2. OS was installed by default with "linux mpath" option. Hardware that triggers the problem: lspci, lspci -n: 03:01.0 Fibre Channel: QLogic Corp. ISP2312-based 2Gb Fibre Channel to PCI-X HBA (rev 02) 03:01.0 0c04: 1077:2312 (rev 02) # cat /proc/scsi/scsi Attached devices: ... Host: scsi1 Channel: 00 Id: 00 Lun: 00 Vendor: COMPAQ Model: HSV111 (C)COMPAQ Rev: 4007 Type: RAID ANSI SCSI revision: 05 ... If /etc/multipath.conf does not contain "hardware_handler" option (by default), multipath consider that hardware handler "hp_sw" should be used: [root@localhost ~]# multipath -v2 -d create: mpath0 (3600508b400011c130003000000fc0000) COMPAQ,HSV111 (C)COMPAQ [size=100G][features=0][hwhandler=1 hp_sw] \_ round-robin 0 [prio=4][undef] \_ 0:0:0:1 sda 8:0 [undef][ready] \_ 0:0:1:1 sdb 8:16 [undef][ready] \_ 1:0:0:1 sdc 8:32 [undef][ready] \_ 1:0:1:1 sdd 8:48 [undef][ready] If we add correct "hardware_handler" option for the device - the problem is gone. So following solves the problem. devices { device { vendor "HP|COMPAQ*" product "HSV1*" hardware_handler "1 hp-sw" } } i think this bug is related to https://bugzilla.redhat.com/show_bug.cgi?id=348951 https://bugzilla.redhat.com/show_bug.cgi?id=226791 it's said the bug is fixed in device-mapper-multipath-0.4.7-17.el5 (in errata), but my problem triggers on device-mapper-multipath-0.4.7-17.el5! [root@localhost ~]# cat /etc/redhat-release CentOS release 5.2 (Final) [root@localhost ~]# rpm -qa |grep mapper device-mapper-event-1.02.24-1.el5 device-mapper-multipath-0.4.7-17.el5 device-mapper-1.02.24-1.el5 ------------------------------------------ Next issue which might be related: mkinitrd (mkinitrd-5.1.19.6-28) script does not put dm-hp-sw.ko module into initrd. I believe this is because of the previous issue. To workaround this i've added following line to the /etc/modprobe.conf: alias scsi_hostadapter2 dm-hp-sw This is incorrect from the logic point of view, but this will give us needed result - dm-hp-sw.ko module will be added into initrd. ------------------------------------------ And the last one: mkinitrd script does not handle "install" primitive for qla2xxx module in /etc/modprobe.conf (which was set by installer!) and thus qla2xxx.ko module is _not_ put into initrd. There is already a bug for it https://bugzilla.redhat.com/show_bug.cgi?id=317811, but it was closed as won't fix.